Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{209056,
        author = {Rudrarajsinh Ghelot and Dr. Swapnesh Taterh},
        title = {Detection Of Malicious Payloads Hidden in Digital Images Using Steganalysis Techniques},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {no},
        pages = {769-779},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=209056},
        abstract = {Digital images are now used everywhere—on social media, in messaging apps, on websites, and on file-sharing services—and this widespread use provides attackers with a convenient way of hiding information. The technique known as steganography, which involves hiding information within media in such a way that its presence is not noticed, has many legitimate applications, but it can just as readily be used to conceal malicious payloads, scripts, or commands within what appears to be a normal picture. This paper looks at the existing literature on image steganalysis with a special emphasis on its application as a security measure to detect malicious payloads, and suggests a security-oriented image analysis system that is intended to identify exactly this type of threat. When images are uploaded, they will be analysed using steganalysis techniques together with other general image features, searching for anomalies that indicate the presence of hidden data; any data that is extracted will then be subject to static analysis to check for harmful content, and each image will be classified as safe, suspicious, or possibly malicious, with the appropriate warning given to the user. The review includes a range of image formats and steganographic methods, with a particular focus on Least Significant Bit (LSB) manipulation, and brings together the statistical, structural, and machine-learning-based detection methods reported in the primary literature examined, including quantitative data on detection rates and accuracy that have been obtained directly from published experiments. The results reported indicate that traditional statistical detectors (RS analysis, chi-square attack, sample pair analysis) are still effective against non-adaptive LSB embedding but perform worse when faced with content-adaptive schemes, whereas convolutional neural network detectors such as Zhu-Net achieve a detection accuracy of about 88% against WOW embedding at 0.4 bits per pixel, representing an improvement of nearly 12 percentage points over the classical SRM-plus-ensemble-classifier baseline. A survey of eighteen documented real-world stegomalware cases also reveals that PNG and JPEG are still the most commonly used carrier formats. The aim is to demonstrate how steganalysis can be used to uncover covert threats that are hidden in images, thus providing a practical security measure before users open or download suspicious files, while also highlighting the areas that still need to be addressed in future work.},
        keywords = {steganography; steganalysis; least significant bit (LSB); stegomalware; digital image forensics; convolutional neural networks; malware detection; image security},
        month = {September},
        }